Reading Room
Practical answers for people who look after historical collections, on capture, text recognition, cataloging, preservation and funding, with sources you can check.
Featured this issue

Featured article, issue:
What digitizing a historical collection actually involves
Why read it: It puts every decision in a digitization project in the order you will meet it, says who should own each one and which standard applies, and links to a deeper guide for every step.
We want to digitize our collection. What happens from first assessment to files online and preserved, what do we have to decide, and what does each step ask of our staff?
25 articles in five sections. The issue date on each article is its month in our editorial calendar; the published date is when it went online.
Capture and imaging
Scanning documents, photographs, negatives, maps and objects to a standard you can check.
- Issue: Digitizing glass plate negatives: handling, capture and filesHow to digitize glass plate negatives safely: handling cracked plates and flaking emulsion, camera or flatbed capture, inversion and which master files to keep.
- Issue: Digitizing oversize maps and drawings: single capture or stitchingHow to digitize large maps, plans and drawings: when sheet-fed scanners are unsafe, single capture or stitching, checking seams, file sizes and naming.
- Issue: Camera or flatbed? Digitizing negatives and slides in an archiveCamera scanning vs flatbed vs film scanner for archive negatives and slides: film bases, inversion, dust, FADGI checks and a decision table by format.
- Issue: Scanning bound volumes without damaging the bindingHow to scan bound books without damaging them: safe opening angles, cradles, overhead capture versus flatbeds, and triage for ledgers, registers and scrapbooks.
- Issue: FADGI star ratings: which level does your collection need?FADGI star ratings grade measured image quality on a target, not a ppi setting. What one to four stars mean, and which level to require for each material.
Text recognition and search
OCR, handwriting recognition, transcription and the search that depends on them.
- Issue: Semantic search for archives: what it adds and what it risksSemantic search for archives explained: how it differs from keyword search, what published tests measured, its risks, and the requirements to set first.
- Issue: Historical newspaper OCR: why it fails and what improves itWhy historical newspaper OCR is so poor, what national programs measured, and how to capture, lay out, re-OCR and correct your own town paper for better search.
- Issue: Volunteers, AI or both? Planning a transcription projectCrowdsourced transcription, machine drafts or both? How to choose for a handwritten collection, redesign volunteer roles, review the text and record its source.
- Issue: OCR or HTR? Matching text recognition to each kind of recordOCR vs HTR for archives: which method fits print, typescripts, forms, ledgers and letters, which output file to ask for, and when typing by hand still wins.
- Issue: Can AI read old handwriting? What HTR and LLMs get rightCan AI read old handwriting? What published tests of HTR engines and language models found, the errors to watch for, and how much human review to plan.
Cataloging and metadata
Describing collections with standards, vocabularies and machine suggestions that people review.
- Issue: Item-level or collection-level description for digitized archivesItem level vs collection level description for digitized archives: when a finding aid with linked images is enough, when items need records, and the minimum fields.
- Issue: Cataloging photographs with unknown people, places and datesCataloging historical photographs when no one knows who, where or when: record the evidence, write uncertain dates consistently and keep AI guesses labeled.
- Issue: Importing digitized records into PastPerfect, CONTENTdm or OmekaPastPerfect import, CONTENTdm batch loading, Omeka CSV Import and CollectiveAccess: how each links files to records, and a delivery spec to follow.
- Issue: How to label AI-generated metadata in catalog recordsHow to label AI-generated metadata field by field in Dublin Core, MARC, EAD and CMS records, with review statuses, rules for edits and a public statement.
- Issue: Dublin Core, MARC or EAD: choosing a metadata standardDublin Core vs MARC vs EAD: start from the system your records live in, learn what each standard is for, and map the dozen fields where most errors happen.
Preservation, rights and access
Keeping files safe for decades, reviewing rights and publishing collections people can use.
- Issue: Digitizing Native American materials: protocols, consent, NAGPRADigitizing Native American archival materials: what the Protocols, the NAGPRA duty of care and Local Contexts ask before capture, not only before publishing.
- Issue: Destroying paper originals after scanning: what state rules allowCan you destroy original records after scanning? How Colorado, Texas, Washington, North Carolina, Oregon and federal rules answer, and which originals to keep.
- Issue: ADA Title II and digitized collections: deadlines and exceptionsDoes the ADA Title II web rule cover your digital collections? The April 2027 and 2028 dates, the archived content exception, and what to fix first.
- Issue: What is IIIF, and does a small museum or archive need it?What is IIIF? Open standards for deep zoom, comparison and image reuse: what they give a small museum or archive, three ways to adopt them, and when to wait.
- Issue: NDSA Levels 2.1: a digital preservation plan for small archivesThe NDSA Levels of Digital Preservation 2.1 applied to a first vendor delivery: copies, checksums, BagIt checks and a fixity schedule a small staff can keep.
Planning, budgets and funding
Assessing a collection, estimating the work, choosing a vendor and paying for it.
- Issue: What digitizing a historical collection actually involvesHow to digitize archives and museum collections: the ten decisions from purpose to preservation, who owns each one, the standards that apply and what staff do.
- Issue: In-house or outsourced digitization: how to decideIn-house vs outsourced digitization compared line by line: equipment, staff time, QC, transport, fees and oversight, with a task-by-task hybrid split.
- Issue: Running a digitization pilot: what to test before you commitHow to run a digitization pilot project: sample your hardest items, set pass criteria first, then measure capture, OCR, metadata, a test import and cost.
- Issue: Writing a digitization RFP that gets comparable bidsHow to write a digitization RFP that yields comparable bids: counts by format, specs by material, a unit-price bid form, QC sampling, pilot and AI clauses.
- Issue: Grants for digitization: NEH, IMLS, NHPRC and Colorado programsDigitization grants as of November 2026: NEH Collections Stewardship is open to December 15, IMLS and NHPRC are closed, and what Colorado programs fund.
The Reading Room is Dastok’s publication for people who look after historical collections and have to make decisions about digitizing them. Each article answers one practical question: what resolution a glass plate negative needs, what a character error rate tells you, how to label a machine-suggested subject, whether a 1912 letter can go online, how to build a grant budget. The five sections above follow the work from capture to funding.
Who writes it
Articles are written and checked by the Dastok Editorial Team. We cite the standard, government page or institution we rely on, give the date we checked facts that change, such as grant deadlines, and label every fictional example as fictional. When an article describes our own method, it says so. Articles are not legal advice; where a question turns on law, they point to the source and suggest who to ask.
How the dates work
Each article carries two dates. The issue date is its month in an editorial calendar that runs from December 2025 to November 2026, which is how the series is arranged by topic and season. It is not a record of company history: Dastok Archive LLC was formed in November 2026. The published date is the day the article actually went online, and an updated date appears when facts are revised after publication.
Suggest a topic
If there is a question you keep running into with your own collection, or a fact in an article that has changed, write to hello@dastok.com with the question and, if it helps, a short description of the material. Suggestions shape what we write next. For help with a specific collection, plan a pilot box or try the planning tools.
Send one box first.
A pilot box shows you the images, the recognized text and the catalog records before you commit a whole collection. Tell us what is in the box and what you want people to find.
Studio and contacts
24 South Weber Street, Suite 225Colorado Springs, CO 80903(719) 244-8450sales@dastok.com